add indexes and columns to support faster selection using state-country information

inputs/modify_tables.sh

+#!/bin/bash
+
+# Assign parameters to variables
+HOST=$1
+USER=$2
+PASSWORD=$3
+
+echo "*** Fix tables to include states and countries data and to be indexed on them:"
+mariadb -h "$HOST" -u "$USER" -p"$PASSWORD"  < inputs/sql/modify_tables_for_state_country.sql

inputs/sql/modify_tables_for_state_country.sql

+-- US data
+-- add index to LODES8 table, query already optimized
+ALTER TABLE LODES8
+ADD INDEX idx_geoid (GeoID);
+-- no need to modify QCEW because there is agg level already to select geographical layer
+
+-- EU data
+-- LFS at the country level only
+-- SBS at the country level only
+
+-- Step 1: Add the country column to the table
+ALTER TABLE REA
+ADD COLUMN eu_country CHAR(2);
+-- Step 2: Add the new column to the table index
+ALTER TABLE REA
+ADD INDEX idx_eu_country (eu_country);
+-- Step 3: Update the country column based on the cntr_code column
+UPDATE REA
+SET eu_country = LEFT(GeoID, 2);

prepare_database.sh

@@ -45,5 +45,8 @@ PASSWORD=$3
 ./inputs/load_eurostat_lfs.sh "$HOST" "$USER" "$PASSWORD"
 ./inputs/load_eurostat_sbs.sh "$HOST" "$USER" "$PASSWORD"
 
+# modify tables to include states and countries data and to be indexed on them
+./inputs/modify_tables.sh "$HOST" "$USER" "$PASSWORD"
+
 # dump data and drop database on 'local' instance
 ./inputs/dump_and_drop.sh "$HOST" "$USER" "$PASSWORD"